Suicide mortality rates are usually second or third after accidents & homicides, depending on the country, among adolescents. Rates in males are regularly higher than those in females. Owing to the universal absence of a systematic recording of parasuicide cases, morbidity data are much more difficult to get & to analyze. Data based on hospital admissions only give approximations. Epidemiologic studies on relapses as well as parasuicide follow-up are rare. Taking into account the number & heterogeneity of known risk factors, suicidal behaviour prediction is based on probability statistics; valuable only for groups, never for individuals. Preventative measures are discussed according to their pertinence in each of the 3 classic levels of prevention. (15 refs.) JA